Nvidia backs OpenAI's 8GW Ohio AI campus
OpenAI, Nvidia and SB Energy have set a 20-year structure for the PORTS-Pike campus. The important constraint is power, financing and construction—not another chip launch.
40 verified stories covering Data Centers, product updates and industry developments.
OpenAI, Nvidia and SB Energy have set a 20-year structure for the PORTS-Pike campus. The important constraint is power, financing and construction—not another chip launch.
Nvidia is working with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR on compute-financing platforms that could mobilize more than $500 billion for AI infrastructure.
Firebird has launched a NVIDIA DSX AI factory in Armenia, with a roadmap for 70,000 GPUs, 300MW in the country, and a wider 2GW frontier-market infrastructure plan.
Volta and Bitdeer disclosed a Norway AI factory deal that Bloomberg-linked reports identify as Anthropic's latest compute push.
Green computing enters China carbon-peak plan is reframed for global technology readers, with the key numbers, product claims and open questions kept intact.
Subsea bandwidth, GPU clusters and AI data-center corridors.
AI power demand, grid queues and private gas-fired capacity.
Omen AI raised a $31 million Series A for inline sensors that track coolant contamination and wear as AI data centers move deeper into liquid cooling.
Valar Atomics powered an Nvidia Blackwell chip from its Ward250 reactor in Utah, framing the test as a first for nuclear-fed AI infrastructure.
Meta is exploring a business that rents out excess AI compute, turning a costly infrastructure buildout into a possible revenue line.
Stathera raised $55 million to replace quartz timing components with smaller silicon oscillators aimed at AI data centers.
South Korea puts $518B behind four AI chip clusters
Upscale AI raised $190 million at a $2 billion valuation, with Nvidia joining the round, as its SkyHammer switch chips pitch open Ethernet standards for large AI clusters.
The U.S. energy regulator ordered six major grid operators to justify or rewrite large-load interconnection rules as AI data centers strain power queues.
Goldman Sachs estimates global AI infrastructure spending from 2026 to 2031 could reach $7.6 trillion, with compute, data centers and power as the major buckets.
The reported 10-gigawatt lease would shift OpenAI from owning more infrastructure to long-term capacity commitments backed by partners such as Nvidia.
China is preparing a five-year AI infrastructure plan that could spend about $295 billion while requiring at least 80% domestic hardware and software.
Meta is partnering with Reliance on a built-to-suit AI data center in Jamnagar, India, starting with 168MW and green power contracts.
Days before SpaceX's planned listing, Elon Musk showed an AI1 satellite design that would move part of AI infrastructure into orbit.
The data-center operator has moved from an $11 billion take-private deal to a potential $50 billion-plus valuation as AI demand turns power and land into scarce assets.
Google says it will replenish more water than its data centers consume by 2030, pairing public water reporting with projects across 97 watersheds.
A Deloitte-SIA report projects AI data-center chip revenue could reach $1.2 trillion by 2028, reframing AI infrastructure as a full-stack semiconductor race.
SoftBank wants to build 5 GW of AI data center capacity in northern France, tying Masayoshi Son’s infrastructure wager to Europe’s nuclear power advantage.
DriveNets raised a $410 million Series D at an $8.5 billion valuation, with AMD joining a bet on Ethernet fabrics for large AI deployments.
Qualcomm used Computex 2026 to introduce Dragonfly, a data center brand covering CPUs, inference accelerators and ASICs, while leaving specifications for later.
DriveNets is valued at $8.5 billion as AMD backs its Ethernet fabric bet against Nvidia-style closed AI clusters.
Anthropic will pay SpaceX $1.25 billion monthly for GPU compute from the Colossus cluster, totaling $45 billion over 36 months.
SendCutSend, a metal fabrication startup, raised $110 million led by Paradigm and Sequoia, hitting a $1 billion valuation as AI data center demand drives orders.
European industrial electricity costs are double those in the US and 50% higher than in China and India, threatening AI data center expansion.
The UK liquid-cooling company wants to scale immersion systems as next-generation AI racks push far beyond what air cooling can handle.
Google’s Project Suncatcher is reportedly moving from research toward launch talks with SpaceX as AI data centers run into power and siting constraints on Earth.
NVIDIA paired a five-year IREN share option with a $3.4 billion GPU cloud commitment, signaling that AI infrastructure now starts with power, land and data centers.
Nvidia is tying itself to Corning as AI data centers move from copper cabling to optical links, funding new U.S. plants and taking warrant-based upside.
AI infrastructure startup Fluidstack, which builds custom data centers for Anthropic, Meta, and others, has more than doubled its valuation to $18 billion within four months, driven by surging demand for specialized AI compute.
On March 1, Iranian Shahed drones struck three Amazon Web Services data centers — two in the UAE and one in Bahrain — in what is considered the first state-backed military strike on commercial cloud infrastructure. The incident exposed critical flaws in multi-AZ redundancy, data localization laws, and the blurring line between military and commercial computing.
AMD's upcoming MI400 series AI accelerators, led by the MI455X with 432GB HBM4 memory and 40 PFLOPS FP4 performance, challenge NVIDIA's dominance on paper, but software ecosystem hurdles remain.
Global data center electricity consumption is projected to exceed 1,000 TWh by the end of 2026, driven almost entirely by AI workloads, raising concerns about grid strain, rising residential bills, and the limits of energy supply.
Nvidia's B200 and GB200 NVL72 racks are sold out until mid-2026, with cloud giants holding over 3.6 million units in backlog. The article examines supply chain bottlenecks, the upcoming Blackwell Ultra (GB300), and what this means for AI companies.
The Trump administration's steep tariffs on imported components are raising data center costs by 20-50%, pushing AI companies to expand in Canada, Nordic countries, and Southeast Asia instead of the United States.
AI model training consumes enormous amounts of electricity, with a single GPT-4-level training run using as much power as hundreds of U.S. homes use in a year. Inference energy may already exceed training, and companies like Microsoft, Google, and DeepSeek are pursuing different strategies to address the growing conflict between AI progress and energy demand.